• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of eukcellPermissionsJournals.ASM.orgJournalEC ArticleJournal InfoAuthorsReviewers
Eukaryot Cell. Dec 2008; 7(12): 2078–2086.
Published online Oct 10, 2008. doi:  10.1128/EC.00128-08
PMCID: PMC2593192

TFIIIC Binding Sites Function as both Heterochromatin Barriers and Chromatin Insulators in Saccharomyces cerevisiae[down-pointing small open triangle]


Chromosomal sites of RNA polymerase III (Pol III) transcription have been demonstrated to have “extratranscriptional” functions, as the assembled Pol III complex can act as chromatin boundaries or pause sites for replication forks, can alter nucleosome positioning or affect transcription of neighboring genes, and can play a role in sister chromatid cohesion. Several studies have demonstrated that assembled Pol III complexes block the propagation of heterochromatin-mediated gene repression. Here we show that in Saccharomyces cerevisiae tRNA genes (tDNAs) and even partially assembled Pol III complexes containing only the transcription factor TFIIIC can exhibit chromatin boundary functions both as heterochromatin barriers and as insulators to gene activation. Both the TRT2 tDNA and the ETC4 site which binds only the TFIIIC complex prevented an upstream activation sequence from activating the GAL promoters in our assay system, effectively acting as chromatin insulators. Additionally, when placed downstream from the heterochromatic HMR locus, ETC4 blocked the ectopic spread of Sir protein-mediated silencing, thus functioning as a barrier to repression. Finally, we show that TRT2 and the ETC6 site upstream of TFC6 in their natural contexts display potential insulator-like functions, and ETC6 may represent a novel case of a Pol III factor directly regulating a Pol II promoter. The results are discussed in the context of how the TFIIIC transcription factor complex may function to demarcate chromosomal domains in yeast and possibly in other eukaryotes.

Eukaryotic genomes are organized into structurally and functionally distinct domains as one layer of transcriptional regulation to allow the expression of particular sets of genes when required and to restrict their expression when necessary. Mechanisms of activation usually involve DNA-bound transcription factors that recruit RNA polymerase or general transcription factors or recruit proteins that promote the formation of chromatin structures compatible with RNA polymerase preinitiation complex formation and transcriptional elongation. Repressive chromatin domains can inhibit gene expression at either of these stages. Chromatin boundary elements function to separate chromosomal domains so that regulatory regions of one domain do not inappropriately influence adjacent domains, either by insulating promoters from activation or by acting as a barrier to propagating repressive heterochromatin (55, 58).

Evidence has accumulated over the past several years that RNA polymerase III (Pol III) promoter sequences, mainly studied using tRNA genes (tDNAs), can possess an intrinsic chromatin boundary activity. This was first demonstrated at the heterochromatic HMR locus in Saccharomyces cerevisiae, as the downstream tDNA is a critical component of the barrier that prevents the inappropriate spreading of silencing from HMR (16), and the characterization of this activity was the first demonstration of a natural chromatin boundary in yeast. Another yeast tDNA, TRT2, was shown to prevent the spread of repression from the MATa cell-specific STE6 gene in MATα cells (51).

In eukaryotes, Pol III is devoted to the transcription of small RNAs participating in basic cellular functions such as protein synthesis (tRNAs, 5S rRNA), pre-mRNA processing (U6 snRNA), and protein secretion (7SL RNA) (19, 46) and has recently been shown to effect micro-RNA expression (5). Additionally, a considerable fraction of the megacopy Alu repetitive elements in primates and B1 elements in mice can contain active Pol III promoters (12, 26, 54). The transcription initiation of 5S rRNAs and tDNAs is dependent on internal control regions (ICRs), which are transcription factor binding sites that lie within the transcribed DNA sequence. Within tDNAs, the ICR is formed by two nonadjacent conserved elements, boxA and boxB. The boxB consensus is conserved in all eukaryotes (GGTTCGANTCC; the underlined C residue is invariant and essential for efficient Pol III complex assembly and transcriptional activity). These ICR elements together form the specific binding site for the multisubunit transcription factor TFIIIC that upon binding to DNA directs the assembly of another multiprotein transcription factor, TFIIIB, to a less conserved region immediately upstream of the transcription start site, which is then followed by the recruitment of Pol III (19, 29). Mutation of the invariant cytosine residue in boxB inactivates both TFIIIC binding to (3) and Pol III transcription of (40) tDNAs.

Pol III-transcribed RNAs are generally very abundant. For example, tRNAs represent approximately 15% of the total RNA of exponentially growing S. cerevisiae cells (57), implying that there are on the order of 3 × 106 tRNA molecules per yeast cell. Thus, each of the 274 nucleus-encoded (24) and additional mitochondrially encoded tDNAs of this organism must be (on average) transcribed on the order of 104 times per generation (or approaching about twice per second, given a 90-min generation time), a value that is considerably higher than the maximal transcription initiation frequency of one initiation every 6 to 8 s estimated for RNA Pol II-transcribed genes in yeast (31). Based on these estimates, one could argue that tDNA and other Pol III promoters are constantly occupied by active transcription complexes.

In addition to active Pol III-transcribed genes, several studies have identified genomic sites that contain partial complexes containing the Pol III transcription factor TFIIIC (and in one case also TFIIIB) but are not occupied by the polymerase itself (21, 37, 42, 47). In S. cerevisiae, these chromosomal locations are called ETC (extra TFIIIC) sites, and in Schizosaccharomyces pombe they are referred to as COC (chromosome-organizing clamp) sites. In S. pombe, particular COC sites act as heterochromatin barriers, but no distinct function was demonstrated for the ETC sites in S. cerevisiae. Interestingly, these nontranscribed TFIIIC binding regions are overrepresented in the intergenic regions of divergently transcribed genes, suggesting a possible function in genome demarcation.

Since tDNAs or simply bound TFIIIC can act as a chromatin barrier element by blocking the spread of heterochromatin, we hypothesized that such chromatin-bound complexes might also function as insulators to gene activation, which would expand the role of the Pol III complex as a chromosomal boundary or landmark element. We show here that in S. cerevisiae both tDNAs and ETC elements can block the activation of genes when juxtaposed between promoters and upstream activation sequences (UAS), in essence functioning as chromatin insulators. We further demonstrate that ETC sites in S. cerevisiae can also function as heterochromatin barriers and that TFIIIC binding in the absence of TFIIIB is sufficient for both insulator and barrier activities. Finally, we identify in yeast TFIIIC binding sites that possess insulator-like functions in their natural contexts, suggesting a broader role for the Pol III complex in regulating Pol II genes and in genome organization.


Strains containing ectopically inserted tDNAs or ETC sites were made by a standard two-step replacement strategy. For modification of the GAL1-10 locus, the URA3 gene was amplified with primers containing homology to the GAL1 and GAL10 coding regions and integrated by homologous recombination to create the gal1-10 intergenicΔ::URA3 strain DDY 2606. Integrants were verified by PCR analysis at both ends of the inserted URA3. Plasmid pDD866, containing 2.4 kb of the GAL1-10 locus (KpnI-SpeI fragment cloned into Bluescript SK+, SGD chromosome II coordinates 277624 to 280057) was constructed, and a BamHI site was introduced at coordinate 278542 between UASG and GAL10 by site-directed mutagenesis to create pDD901. Wild-type and boxB mutant tDNAs and ETC sites were cloned into this BamHI site, resulting plasmids were linearized and transformed into DDY 2606, and 5-fluorootic acid-resistant colonies were isolated. Mutant boxB sequences were created using site-directed mutagenesis to change the invariant cytosine residue to guanine (boxB consensus, GGTTCGANTCC [invariant C underlined]). This mutation inactivates Pol III genes by inhibiting TFIIIC binding. The resulting isolates were confirmed by both PCR and DNA sequencing to verify proper integration of the TFIIIC binding sites at GAL1-10. Insertion of TRT2 between UASG and GAL1 was performed in a similar manner, by inserting the BamHI site at coordinate 278710 (pDD872). The TRT2 fragment contained sequences between chromosome XI coordinates 46730 and 46826, and the ETC4 fragment spanned chromosome VII coordinates 1010900 to 1010990.

Strains containing the ETC4 site adjacent to the silenced HMR locus were constructed in a similar fashion. DDY 811 and DDY 814 were described previously (32). Plasmid pDD662 contains a 2.5-kb ADE2 fragment (chromosome XV coordinates 566829 to 564325) inserted downstream of HMR (SacI-SalI HMR fragment) at SGD chromosome III coordinate 295736, with HMR-tDNA deleted and replaced with an SphI site. Wild-type and boxB mutant ETC sites were cloned into the SphI-cut plasmid, and the resulting plasmids were linearized and integrated into an hmrΔ::URA3 sir4Δ strain (DDY 631). Ade-positive isolates were screened for a nonmating phenotype, 5-fluorootic acid resistance, and uracil auxotrophy to indicate the integration of the HMR fragment and were confirmed for proper integration by PCR analysis. Positive integrants were then transformed with a SIR4-expressing plasmid and crossed to DDY 511, and resulting diploids were sporulated to obtain SIR-positive isolates containing the HMR-ADE2 constructs with ETC4 inserted in place of HMR-tDNA. Strains containing modifications at the STE6-CBT1 locus were made as described previously (51). All chromosomal coordinates were derived from the Saccharomyces Genome Database (www.yeastgenome.org).

Chromatin immunoprecipitation was performed as described previously (48) using Chelex-100 modification (39) to deproteinize the final DNA. Primer sequences used for PCR analysis are available on request. Anti-FLAG epitope antibody was purchased form Sigma Chemical Co. (Anti-FLAG-M2, catalog number F3165). Northern blots were performed as described by Simms et al. (51), and each was analyzed on a Molecular Dynamics PhosphorImager. Quantitation of mRNA levels of each mutant was performed three times using at least two independent isolates, and quantitative results represent all replicates performed, with representative blot images shown. Growth on galactose was on yeast nitrogen base minimal medium (catalog number Y2025; U.S. Biologicals) containing 2% galactose as the sole carbon source and supplemented to cover all other auxotrophies. Epigenetic silencing of ADE2 was assayed as described previously (32).


Insertion of a tDNA into an ectopic site creates an insulator to gene activation.

In order to directly test the hypothesis that tDNAs can function as insulators, we inserted wild-type and mutant versions of the TRT2 tDNA into the GAL1-10 locus to ask whether this tDNA could block the activation of GAL gene expression by the well-characterized UASG. We have previously characterized a barrier-like activity for this tDNA both ectopically (16) and in its native location (51). TRT2 fragments were cloned between UASG and GAL10 or GAL1 (Fig. (Fig.1A)1A) and then integrated into the gal1-10 intergenic Δ::URA3 strain (DDY 2606) (see Materials and Methods). Resulting strains were grown in raffinose, and GAL transcription was induced by the addition of galactose to a final concentration of 2%. Total RNA was isolated and analyzed by Northern blot analysis using GAL10 or GAL1 probes, and growth on galactose plates was assessed.

FIG. 1.
Insulator activity of a tDNA inserted at the GAL1-10 locus. Insertion of a functional tDNA between UASG and GAL10 (at A) or GAL1 (at B) promoters can insulate the promoter from activation. (A) Schematic diagram of modified GAL1-10 loci. Functional and ...

When TRT2 was inserted in either orientation between UASG and GAL10, the ability to grow on galactose as a sole carbon source was completely abolished, suggesting that TRT2 insulated the GAL10 promoter from the UAS (Fig. (Fig.1B,1B, compare wedges 1 and 2 to wedges 3, 4, 11, and 12). The insertion of an inactive trt2 gene with a boxB point mutation abolished this insulator effect, as indicated by normal growth on galactose (wedges 5, 6, 11, and 12), while maintaining the same UAS-GAL10 promoter spacing as in the insulated strains. The insertion of TRT2 on the GAL1 side of the UAS also prevented growth on galactose, presumably by insulating GAL1 from the UAS (wedges 7, 8, and 15), while the mutant trt2 did not prevent growth (wedges 9, 10, and 16).

Northern blot analysis of GAL gene expression of these strains confirmed an insulator-like effect. When wild-type TRT2 was inserted between GAL1 and UASG, galactose-induced GAL1 transcripts were practically undetectable, while this insertion had no effect on GAL10 (compare lanes 1 and 2 to lanes 7 and 8 in Fig. Fig.1C1C for both GAL10 and GAL1 probes). This is a defining characteristic of an insulator, in that it blocks the activation of a gene only when placed between the transcription factor binding site and the promoter. A slightly different result was seen when TRT2 was placed between GAL10 and the UAS, as both genes were inhibited. GAL10 transcripts were undetectable (Fig. (Fig.1C,1C, lanes 3 and 4), but GAL1 was also reduced. The insertion of an inactive trt2 boxB mutant sequence had a minimal effect on the level of GAL10 transcripts (Fig. (Fig.1C,1C, lanes 5 and 6) and did not affect growth on galactose (Fig. (Fig.1B).1B). These results demonstrate that a functional tDNA positioned between a UAS and promoter in yeast has the potential to behave as a typical eukaryotic insulator, blocking enhancer/UAS promoter communication when placed between the two elements. Since the insertion of TRT2 on the GAL1 side had no effect on GAL10 expression, the assembled Pol III complex did not appear to be sterically interfering with the binding of Gal4p to the UAS in this case, or transcript levels of both genes would have been reduced.

ETC sites can function both as an insulator to gene expression and as a barrier to repression.

Although ETC sites have been identified in several studies as sites that bind TFIIIC but not the rest of the Pol III complex (21, 37, 47), little has been done to identify potential functions for the ETC, or extra TFIIIC, sites. Interestingly, most of the ETC sites identified lie between divergently transcribed genes. We hypothesized that bound TFIIIC alone might be able to function as an insulator between regulatory elements of divergently transcribed promoters.

In order to test the hypothesis that ETC sites can function as insulators, we cloned a 90-bp fragment of the ETC4 site between the GAL10 gene and UASG as described for Fig. Fig.11 above. Strains were constructed containing both wild-type ETC4 and etc4 boxB mutant sequences in both orientations and integrated into the GAL locus. Independent colony isolates were streaked onto minimal medium containing galactose as the sole carbon source. The results in Fig. Fig.22 indicate that when the wild-type ETC4 site is inserted between the UAS and the promoter, growth on galactose is impaired (Fig. (Fig.2A,2A, wedges 3 to 6, compared to wild-type growth in wedges 1 and 2), indicating that the GAL10 gene is again insulated from the UAS, presumably due to the binding of the TFIIIC complex to the ETC4 sequence. However, when the boxB mutant etc4 site (presumably unable to bind TFIIIC) was inserted at the same location (Fig. (Fig.2A,2A, wedges 8 to 11), the cells retain their ability to grow on galactose. Deletion of just the UASG also prevents growth on galactose, as expected (Fig. (Fig.2A,2A, wedge 12). Northern blot analysis of GAL10 transcripts shows a slightly different result than with TRT2, as the wild-type ETC4 sequence does not completely inhibit GAL10 induction (Fig. (Fig.2B).2B). This is most likely due to the fact that TRT2 recruits the entire Pol III complex, while ETC sites recruit only TFIIIC (see Fig. Fig.44 below and Discussion).

FIG. 2.
A single ETC site can insulate the GAL10 promoter from UASG. (A) Wild-type and mutant ETC4 sequences were inserted between GAL10 and UASG as described for Fig. Fig.1.1. Wedges: 1 and 2, DDY 3 and DDY 2864, wild-type (wt) GAL locus; 3 and 4, DDY ...
FIG. 4.
Chromatin immunoprecipitation analysis of Pol III factors at tDNA and ETC4 boundaries. Strains containing TRT2 or ETC4 at the GAL locus (A) or the native tDNA or ETC4 at HMR (B) and expressing the FLAG epitope attached to either the TFIIIC subunit Tfc1p ...

Since tDNAs can function both as barriers to repression and as insulators to activation, we next asked whether ETC4 could also act as a barrier to Sir protein-mediated silencing. We have previously constructed a reporter system that contains the ADE2 gene cloned downstream from the HMR silent mating locus (32), which grows as pure white Ade-positive colonies on minimal medium containing limiting adenine (Fig. (Fig.3A).3A). Deletion of HMR-tDNA from this region results in an epigenetic spread of silencing, yielding a mixture of white (unsilenced), red-to-pink (partially silenced), and white-red sectored colonies (Fig. (Fig.3B).3B). Removal of the tDNA therefore partially weakens the HMR barrier so that silencing ectopically spreads in a variegated fashion. Replacement of the tDNA with ETC4 in either orientation restores the white-colony phenotype (Fig. 3C and D), suggesting that bound TFIIIC alone is sufficient to stabilize the barrier. Replacement with a mutant etc4 containing the boxB-inactivating mutation yields a mixed-colony-color phenotype similar to that of tdnaΔ strains (Fig. 3E and F). The silencing of ADE2 in the etc4 boxB strains is Sir dependent, as subsequent deletion of SIR4 results in white colonies (Fig. 3G and H). We have previously confirmed that the silencing of ADE2 in this system is due to heterochromatin spreading from HMR, as strains deleted for the HMR silencers also yield pure white colonies (32).

FIG. 3.
The ETC4 sequence can function as a barrier to heterochromatin spreading. Strains containing the ADE2 gene recombined downstream of HMR were used to assess ectopic spreading of silencing, as indicated by the formation of pink-to-red colonies on minimal ...

TFIIIC binding but not TFIIIB binding is required for boundary activity of ETC4.

To this point, we have made two assumptions regarding the ETC4 site in both our insulator and barrier assays. The first is that when moved to an ectopic location, ETC4 behaves as in its native location, in that it binds TFIIIC but not TFIIIB. Second, we have assumed that the mutation of the conserved cytosine in boxB inhibits TFIIIC binding as completely as it does in tDNAs (3). In order to confirm these assumptions and the role of TFIIIC in creating boundaries, we crossed our boundary reporter strains with strains engineered to have the FLAG epitope attached to Tfc1p, the Tau95 subunit of TFIIIC, and to FLAG-tagged Brf1p strains, marking the 70-kDa subunit of TFIIIB. The resulting strains were analyzed by chromatin immunoprecipitation using anti-FLAG antibody. Figure Figure4A4A shows the expected enrichment over background (no antibody controls) for both Tfc1p and Brf1p at the TRT2-GAL locus. When TRT2 was replaced with ETC4, Tfc1p enrichment was observed, but Brf1p enrichment was not, confirming that the ETC4 site binds only TFIIIC but not TFIIIB when moved to an ectopic location. Replacing with the boxB mutant etc4 site, which lacks insulating activity, showed no enrichment for either tagged protein compared to the no-antibody control, confirming that the point mutation inhibits TFIIIC binding as predicted and correlating TFIIIC binding with insulator activity. A distal control tDNA showed equal enrichment in each strain. Similar results were obtained for the comparable HMR-ADE2 ETC4 barrier strains, shown in Fig. Fig.4B.4B. These results demonstrate that TFIIIC binding alone, in the absence of TFIIIB, is associated with both barrier and insulator activities in these strains.

tDNAs and ETC sites can exhibit insulator-like properties in their natural context.

We next asked whether Pol III binding sites in their natural contexts actually exhibit insulator-like properties. To address this possibility, we revisited the STE6-CBT1 locus. Our previous work demonstrated that TRT2, a tDNA that lies between divergently transcribed STE6 and CBT1, exerts an apparent inhibitory position effect on CBT1 in MATa cells, where STE6 is active (51). Complete deletion or boxB point mutation of TRT2 resulted in an increase in CBT1 expression in MATa cells. Current interpretations of tDNA position effects on Pol III transcription have suggested that the presence of the Pol III complex has a negative influence on neighboring Pol II promoters, hypothesized to be due to factors such as nucleolar localization or nucleosome-positioning effects (4, 33, 56). We reasoned that an alternative hypothesis, an insulator-like activity of assembled Pol III complexes, could also explain the increased CBT1 expression observed for MATa trt2Δ strains. In MATa cells, the transcription factor Mcm1p binds to a sequence within the α2 operator to activate the transcription of STE6 (18). We hypothesized that in the absence of a functioning tDNA and assembled Pol III complex, the increase in CBT1 expression upon TRT2 mutation could be due to inappropriate activation of CBT1 by α2 operator-bound Mcm1p in MATa cells.

To ask if this position effect was due to the insulator function of TRT2, we created mutant yeast strains with TRT2 either deleted or containing the boxB mutation to inactivate Pol III complex assembly and TRT2 transcription. We then constructed yeast strains that additionally contained a deletion within the α2 operator, specifically the Mcm1p binding site within the STE6 UAS. An analysis of CBT1 expression in these strains is shown in Fig. Fig.55.

FIG. 5.
TRT2 functions as an insulator in its natural context. (A) Northern blot analysis of CBT1 expression in MATa strains containing mutations in the TRT2 tDNA and STE6 regulatory elements. Mutation of TRT2 results in an ~2.4-fold increase in CBT1 ...

Deletion or mutation of TRT2 resulted in an increase in CBT1 expression (compare Fig. Fig.5A,5A, lanes 3 to 6 with lanes 1 and 2; quantitation in Fig. Fig.5B)5B) characteristic of a tDNA position effect. However, when either the entire STE6 UAS or just the α2 operator site (the Mcm1p binding site) was also deleted, the increase in transcription of CBT1 was partially mitigated. These results indicated that at least part of the increased expression due to tDNA deletion was due to the inappropriate activation of CBT1 by transcription factors (Mcm1p) bound to the STE6 UAS and not solely due to factors such as subnuclear localization or nucleosome positioning exerting a direct negative effect on the neighboring Pol II promoter.

We conducted a similar analysis on the ETC6 site, which lies between TFC6 and ESC2 on chromosome IV. This was an interesting locus to study; as TFC6 encodes a component of the TFIIIC complex, it has been suggested that ETC6 may mediate the autoregulation of TFC6 (37). As shown in Fig. Fig.6,6, chromosomal deletion of ETC6 results in a subtle yet reproducible 2.4-fold increase in TFC6 transcripts (confirmed with six independent isolates). To test if this deletion allowed an uncharacterized upstream UAS to activate TFC6, we made strains deleted for both ETC6 and the entire upstream region back to the ESC2 start codon. These strains showed a marked decrease in TFC6 transcripts (Fig. (Fig.6),6), which resulted in a slow-growth phenotype that yielded colonies of variegated size (D. Donze, unpublished observations), presumably due to limiting TFC6 synthesis. The increase in TFC6 transcripts upon deletion of ETC6 coupled with a subsequent decrease upon the deletion of upstream sequences is consistent with an insulator-like function of ETC6.

FIG. 6.
ETC6 displays an insulator-like activity in its natural context. (A) Strains were created that contained chromosomal deletions of either the ETC6 site boxB sequence or both ETC6 and the upstream region. Deletion of ETC6 resulted in an ~2.5-fold ...


A growing body of evidence suggests that genes transcribed by RNA Pol III, mainly the tDNAs, can be involved in various extratranscriptional functions throughout eukaryotic genomes (10, 51). Mostly studied with the yeast Saccharomyces cerevisiae, these additional functions include directing the periodic integration of Ty elements (1, 9, 14, 35) and the integration of a Dictyostelium retrotransposon (52), creating pause sites for replication fork progression (13), the dominant overriding of nucleosome positioning sequences (38), and creating the apparent repression of transcription from neighboring RNA Pol II promoters (4, 30, 34, 51) and, conversely, the protection of neighboring Pol II genes from transcriptional repression due to propagating heterochromatin structures or the effects of other transcriptional repressors (16, 42, 45, 49, 51). Most recently, a role for the Pol III complex in sister chromatid cohesion has been demonstrated (17), as well as a role in recruiting the condensin complex to chromosomes (11, 23). Some of these genomic effects were believed to require a fully functional RNA Pol III complex bound to a tDNA, but recent data indicate that partial Pol III complexes bound to DNA are in some cases sufficient to impart certain extratranscriptional activities (37, 42). Given that S. cerevisiae contains 274 tDNAs scattered throughout its genome (22), and vertebrates contain in addition to tDNAs many repetitive elements capable of recruiting the Pol III apparatus (54), such extratranscriptional effects may exert a substantial effect on genome-wide chromosomal organization in eukaryotes. This study expands the role of tDNAs and ETC/COC sites as potential chromatin-organizing elements, demonstrating an additional insulator activity of DNA-bound Pol III complexes.

One important question regarding Pol III boundary activity centers on which components of the Pol III complex are required for boundary function. Compared to previous studies, the results presented here suggest that the requirement is highly context dependent. Earlier studies on the HMR-tDNA barrier suggested that both TFIIIC and TFIIIB binding, and possibly transcription by Pol III itself, were required to block silencing (16), and a tDNA heterochromatin barrier in S. pombe requires a fully assembled Pol III complex (50). However, the discovery of the heterochromatin barrier function of COC sites in S. pombe (42) challenges this requirement, as only the TFIIIC complex is bound to these sites. We show here that a single ETC site, confirmed to bind only TFIIIC and not TFIIIB (and therefore presumably not Pol III), can prevent the ectopic spread of silencing from the HMR locus. One key difference between these conflicting results is that in the previous study (16), the tDNA barrier was moved in between the HMR-E silencer and the a1 gene and then the a1 gene was used as the reporter gene. It has long been known that the HMR-E silencer is more robust and independent than the HMR-I silencer (7). In this study, we placed putative barrier elements in the natural location downstream from HMR-I. In this downstream location, TFIIIC binding is sufficient to stabilize a barrier that prevents the spread of Sir protein-mediated silencing. We have cloned the ETC4 site between HMR-E and a1 and have found that it functions only as a weak barrier to heterochromatin spreading when close to HMR-E (D. Donze, unpublished data), further demonstrating the context dependence of barrier complexes.

We also present data that TFIIIC binding sites can have a newly identified insulator function, as either a tDNA or an ETC sequence can block the interaction of Gal4p with the GAL10 promoter. Chromatin immunoprecipitation analysis again confirmed that at the ectopic ETC site, only TFIIIC and not TFIIIB are bound in this assay system. We also demonstrated that the TRT2 tDNA in its natural context serves as an insulator between the STE6 and CBT1 genes, preventing the STE6 regulatory elements from affecting CBT1 transcription levels. A similar insulating effect, or to use another electrical circuit analogy, a “resistor-like” effect, is conferred by the ETC6 site in the TFC6 promoter. This result is particularly interesting, as it suggests that the binding of TFIIIC to ETC6 may directly modulate TFC6 expression, providing a potential feedback inhibition by a component of the TFIIIC complex. A detailed analysis of the TFC6 promoter and regulatory elements will be required to verify this hypothesis. These are the first demonstrations of natural insulator-like activities in budding yeast, and the binding of TFIIIC to ETC6 may represent the first example of a Pol III transcription factor directly regulating a Pol II promoter.

With the discovery of the ETC and COC loci in yeast, a key question is whether these TFIIIC binding sites are bona fide regulatory elements. The fact that TFIIIC-only binding sites exist in multiple organisms, coupled with the fact that the ETC loci are conserved among the budding yeast (25, 37, 47), suggests a conserved evolutionary function. It is interesting to note that in the study by Noma et al. (42), most of the COC sites in S. pombe lie between divergently transcribed genes, and inspection of the ETC-like sites reported in the three independent studies with S. cerevisiae reveals that six out of eight of these TFIIIC-bound boxB sequences also lie between divergently transcribed promoters. Additionally, one site in S. cerevisiae (iYGR033C) has been identified that appears to bind TFIIIC and TFIIIB but not Pol III (21), and this site also lies between the divergent promoters of TIM21 and RPL26B. Given this propensity, we propose the hypothesis that these TFIIIC binding sites function as insulators between such divergent promoters in the compact yeast genome, and we are systematically mutating each of these boxB sites to test for effects on the transcription levels of the neighboring Pol II genes. While the number of ETC sites in S. cerevisiae are certainly limited, our data presented here suggest that other tDNAs also may serve as insulators between divergently transcribed genes in yeast, and such effects may be more widespread in S. pombe and other eukaryotes.

Another unresolved question is how DNA-bound TFIIIC functions as a boundary element. Our results here show that recruitment of the entire Pol III complex is not always necessary for this activity. Numerous potential mechanisms exist, some of which are subnuclear localization into a Pol II-depleted nuclear region (56), nucleosome displacement and stable occupancy by TFIIIC (16, 43), and also recruitment of chromatin remodeling complexes. With regard to barrier function, purified human TFIIIC complex has been shown to possess intrinsic acetyltransferase activity (27), but this activity has not been demonstrated in yeast. However, the chromatin remodeling complex ISW2 is recruited to tDNA loci via an interaction with the TFIIIB subunit Bdp1p (1, 20), suggesting a possible role for nucleosome sliding or displacement in tDNA boundary function. However, since TFIIIB is not recruited to ETC sites, it is unlikely that ISW2 is required for the ETC site boundary activity reported in this study.

Our results clearly demonstrate that TFIIIC binding in the absence of the rest of the Pol III machinery is capable of establishing a chromatin boundary in certain contexts, but interaction with additional chromatin modifiers that interact with TFIIIB and Pol III may assist or stabilize the formation of boundaries. Interestingly, the loss of RSC (remodels the structure of chromatin) complex function results in changes in both nucleosome positioning and density near tDNAs (44). Our previous genetic analysis of the HMR-tDNA boundary (32) demonstrated a variegated loss of barrier activity upon mutation of the RSC2 gene, suggesting that chromatin remodeling by RSC also contributes to barrier formation. The RSC chromatin remodeling complex has been shown to be directly recruited to loci transcribed by Pol III (41), and this interaction appears to involve a direct interaction with the Rpb5p subunit conserved among all three eukaryotic RNA polymerases (53). Further complicating any interpretation of the role of the RSC complex are the observations that RSC mutations compromise the recruitment of the cohesin complex to chromosomal arms (2, 28) and that cohesin mutants are defective in barrier activity at HMR (15).

Recent studies have further implicated the Pol III machinery in the recruitment of the condensin class of chromosome binding and organizing proteins. Multiple studies have shown by chromatin immunoprecipitation that the condensin subunits associate with Pol III genes and TFIIIC-only binding sites and that a direct interaction occurs between condensin subunits and TFIIIB or TFIIIC (11, 23).

As for insulator function in the compact yeast genome, the mechanisms involved will likely be different from those proposed for metazoan systems, which can involve long-range chromosomal looping (36, 55). In yeast, it is likely that the Pol III complex or TFIIIC alone may simply physically block the assembly of complexes connecting the UAS to the promoter. For example, simply tethering the Escherichia coli LexA protein to the GAL locus can block Gal4p-mediated activation of a GAL1-lacZ fusion (8), suggesting that a simple physical presence may be sufficient to disrupt coactivator recruitment or perhaps block the propagation of histone acetylation, which has been demonstrated to occur in yeast from sites of targeted acetyltransferase binding (59).

However, chromatin boundary formation by Pol III transcription factor binding sites is turning out to be an extremely complex process, which appears to involve several active complexes such as RSC, and possibly ISW2 and direct histone acetyltransferase recruitment. Boundary function also involves direct structural components such as TFIIIC binding and the involvement of other chromatin architectural components such as bromodomain proteins (32), Nhp6 proteins (6), and, as described above, condensins and cohesions. Future work will need to focus on dissecting how these various DNA and chromatin interacting factors cooperate to create a stable chromatin boundary and to investigate the likely possibility that like promoters and enhancers, different boundaries will utilize different complements of DNA-bound and chromatin-associated factors. Also, continued genome-wide analysis of Pol III factor-mediated boundary activity will be required to determine the overall scope of these effects along eukaryotic genomes. Finally, the extent of Pol III-mediated chromosomal position effects in other eukaryotes could be far reaching, as in human chromosomes a large number of potential TFIIIC binding boxB sequences exist within the repetitive megacopy Alu elements.

S. cerevisiae strains used and generated in this studya


We sincerely thank Giorgio Dieci and Marc Gartenberg for critical reading of the manuscript and many constructive comments.

This work was funded by grants from the National Science Foundation (MCB-0342113 and MCB-0817823 to D.D.).


[down-pointing small open triangle]Published ahead of print on 10 October 2008.


1. Bachman, N., M. E. Gelbart, T. Tsukiyama, and J. D. Boeke. 2005. TFIIIB subunit Bdp1p is required for periodic integration of the Ty1 retrotransposon and targeting of Isw2p to S. cerevisiae tDNAs. Genes Dev. 19955-964. [PMC free article] [PubMed]
2. Baetz, K. K., N. J. Krogan, A. Emili, J. Greenblatt, and P. Hieter. 2004. The ctf13-30/CTF13 genomic haploinsufficiency modifier screen identifies the yeast chromatin remodeling complex RSC, which is required for the establishment of sister chromatid cohesion. Mol. Cell. Biol. 241232-1244. [PMC free article] [PubMed]
3. Baker, R. E., O. Gabrielsen, and B. D. Hall. 1986. Effects of tRNATyr point mutations on the binding of yeast RNA polymerase III transcription factor C. J. Biol. Chem. 2615275-5282. [PubMed]
4. Bolton, E. C., and J. D. Boeke. 2003. Transcriptional interactions between yeast tRNA genes, flanking genes and Ty elements: a genomic point of view. Genome Res. 13254-263. [PMC free article] [PubMed]
5. Borchert, G. M., W. Lanier, and B. L. Davidson. 2006. RNA polymerase III transcribes human microRNAs. Nat. Struct. Mol. Biol. 131097-1101. [PubMed]
6. Braglia, P., S. L. Dugas, D. Donze, and G. Dieci. 2007. Requirement of Nhp6 proteins for transcription of a subset of tRNA genes and heterochromatin barrier function in Saccharomyces cerevisiae. Mol. Cell. Biol. 271545-1557. [PMC free article] [PubMed]
7. Brand, A. H., L. Breeden, J. Abraham, R. Sternglanz, and K. Nasmyth. 1985. Characterization of a “silencer” in yeast: a DNA sequence with properties opposite to those of a transcriptional enhancer. Cell 4141-48. [PubMed]
8. Brent, R., and M. Ptashne. 1984. A bacterial repressor protein or a yeast transcriptional terminator can block upstream activation of a yeast gene. Nature 312612-615. [PubMed]
9. Chalker, D. L., and S. B. Sandmeyer. 1992. Ty3 integrates within the region of RNA polymerase III transcription initiation. Genes Dev. 6117-128. [PubMed]
10. Conesa, C., R. Ruotolo, P. Soularue, T. A. Simms, D. Donze, A. Sentenac, and G. Dieci. 2005. Modulation of yeast genome expression in response to defective RNA polymerase III-dependent transcription. Mol. Cell. Biol. 258631-8642. [PMC free article] [PubMed]
11. D'Ambrosio, C., C. K. Schmidt, Y. Katou, G. Kelly, T. Itoh, K. Shirahige, and F. Uhlmann. 2008. Identification of cis-acting sites for condensin loading onto budding yeast chromosomes. Genes Dev. 222215-2227. [PMC free article] [PubMed]
12. Deininger, P. L., and M. A. Batzer. 2002. Mammalian retroelements. Genome Res. 121455-1465. [PubMed]
13. Deshpande, A. M., and C. S. Newlon. 1996. DNA replication fork pause sites dependent on transcription. Science 2721030-1033. [PubMed]
14. Devine, S. E., and J. D. Boeke. 1996. Integration of the yeast retrotransposon Ty1 is targeted to regions upstream of genes transcribed by RNA polymerase III. Genes Dev. 10620-633. [PubMed]
15. Donze, D., C. R. Adams, J. Rine, and R. T. Kamakaka. 1999. The boundaries of the silenced HMR domain in Saccharomyces cerevisiae. Genes Dev. 13698-708. [PMC free article] [PubMed]
16. Donze, D., and R. T. Kamakaka. 2001. RNA polymerase III and RNA polymerase II promoter complexes are heterochromatin barriers in Saccharomyces cerevisiae. EMBO J. 20520-531. [PMC free article] [PubMed]
17. Dubey, R. N., and M. R. Gartenberg. 2007. A tDNA establishes cohesion of a neighboring silent chromatin domain. Genes Dev. 212150-2160. [PMC free article] [PubMed]
18. Gavin, I. M., M. P. Kladde, and R. T. Simpson. 2000. Tup1p represses Mcm1p transcriptional activation and chromatin remodeling of an a-cell-specific gene. EMBO J. 195875-5883. [PMC free article] [PubMed]
19. Geiduschek, E. P., and G. A. Kassavetis. 2001. The RNA polymerase III transcription apparatus. J. Mol. Biol. 3101-26. [PubMed]
20. Gelbart, M. E., N. Bachman, J. Delrow, J. D. Boeke, and T. Tsukiyama. 2005. Genome-wide identification of Isw2 chromatin-remodeling targets by localization of a catalytically inactive mutant. Genes Dev. 19942-954. [PMC free article] [PubMed]
21. Guffanti, E., R. Percudani, O. Harismendy, J. Soutourina, M. Werner, M. G. Iacovella, R. Negri, and G. Dieci. 2006. Nucleosome depletion activates poised RNA polymerase III at unconventional transcription sites in Saccharomyces cerevisiae. J. Biol. Chem. 28129155-29164. [PubMed]
22. Guldener, U., M. Munsterkotter, G. Kastenmuller, N. Strack, J. van Helden, C. Lemer, J. Richelles, S. J. Wodak, J. Garcia-Martinez, J. E. Perez-Ortin, H. Michael, A. Kaps, E. Talla, B. Dujon, B. Andre, J. L. Souciet, J. De Montigny, E. Bon, C. Gaillardin, and H. W. Mewes. 2005. CYGD: the Comprehensive Yeast Genome Database. Nucleic Acids Res. 33D364-D368. [PMC free article] [PubMed]
23. Haeusler, R. A., M. Pratt-Hyatt, P. D. Good, T. A. Gipson, and D. R. Engelke. 2008. Clustering of yeast tRNA genes is mediated by specific association of condensin with tRNA gene transcription complexes. Genes Dev. 222204-2214. [PMC free article] [PubMed]
24. Hani, J., and H. Feldmann. 1998. tRNA genes and retroelements in the yeast genome. Nucleic Acids Res. 26689-696. [PMC free article] [PubMed]
25. Harismendy, O., C. G. Gendrel, P. Soularue, X. Gidrol, A. Sentenac, M. Werner, and O. Lefebvre. 2003. Genome-wide location of yeast RNA polymerase III transcription machinery. EMBO J. 224738-4747. [PMC free article] [PubMed]
26. Hasler, J., and K. Strub. 2006. Alu elements as regulators of gene expression. Nucleic Acids Res. 345491-5497. [PMC free article] [PubMed]
27. Hsieh, Y. J., T. K. Kundu, Z. Wang, R. Kovelman, and R. G. Roeder. 1999. The TFIIIC90 subunit of TFIIIC interacts with multiple components of the RNA polymerase III machinery and contains a histone-specific acetyltransferase activity. Mol. Cell. Biol. 197697-7704. [PMC free article] [PubMed]
28. Huang, J., J. M. Hsu, and B. C. Laurent. 2004. The RSC nucleosome-remodeling complex is required for Cohesin's association with chromosome arms. Mol. Cell 13739-750. [PubMed]
29. Huang, Y., and R. J. Maraia. 2001. Comparison of the RNA polymerase III transcription machinery in Schizosaccharomyces pombe, Saccharomyces cerevisiae and human. Nucleic Acids Res. 292675-2690. [PMC free article] [PubMed]
30. Hull, M. W., J. Erickson, M. Johnston, and D. R. Engelke. 1994. tRNA genes as transcriptional repressor elements. Mol. Cell. Biol. 141266-1277. [PMC free article] [PubMed]
31. Iyer, V., and K. Struhl. 1996. Absolute mRNA levels and transcriptional initiation rates in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 935208-5212. [PMC free article] [PubMed]
32. Jambunathan, N., A. W. Martinez, E. C. Robert, N. B. Agochukwu, M. E. Ibos, S. L. Dugas, and D. Donze. 2005. Multiple bromodomain genes are involved in restricting the spread of heterochromatic silencing at the Saccharomyces cerevisiae HMR-tRNA boundary. Genetics 171913-922. [PMC free article] [PubMed]
33. Kendall, A., M. W. Hull, E. Bertrand, P. D. Good, R. H. Singer, and D. R. Engelke. 2000. A CBF5 mutation that disrupts nucleolar localization of early tRNA biosynthesis in yeast also suppresses tRNA gene-mediated transcriptional silencing. Proc. Natl. Acad. Sci. USA 9713108-13113. [PMC free article] [PubMed]
34. Kinsey, P. T., and S. B. Sandmeyer. 1991. Adjacent pol II and pol III promoters: transcription of the yeast retrotransposon Ty3 and a target tRNA gene. Nucleic Acids Res. 191317-1324. [PMC free article] [PubMed]
35. Kirchner, J., C. M. Connolly, and S. B. Sandmeyer. 1995. Requirement of RNA polymerase III transcription factors for in vitro position-specific integration of a retroviruslike element. Science 2671488-1491. [PubMed]
36. Kuhn, E. J., and P. K. Geyer. 2003. Genomic insulators: connecting properties to mechanism. Curr. Opin. Cell Biol. 15259-265. [PubMed]
37. Moqtaderi, Z., and K. Struhl. 2004. Genome-wide occupancy profile of the RNA polymerase III machinery in Saccharomyces cerevisiae reveals loci with incomplete transcription complexes. Mol. Cell. Biol. 244118-4127. [PMC free article] [PubMed]
38. Morse, R. H., S. Y. Roth, and R. T. Simpson. 1992. A transcriptionally active tRNA gene interferes with nucleosome positioning in vivo. Mol. Cell. Biol. 124015-4025. [PMC free article] [PubMed]
39. Nelson, J. D., O. Denisenko, P. Sova, and K. Bomsztyk. 2006. Fast chromatin immunoprecipitation assay. Nucleic Acids Res. 34e2. [PMC free article] [PubMed]
40. Newman, A. J., R. C. Ogden, and J. Abelson. 1983. tRNA gene transcription in yeast: effects of specified base substitutions in the intragenic promoter. Cell 35117-125. [PubMed]
41. Ng, H. H., F. Robert, R. A. Young, and K. Struhl. 2002. Genome-wide location and regulated recruitment of the RSC nucleosome-remodeling complex. Genes Dev. 16806-819. [PMC free article] [PubMed]
42. Noma, K., H. P. Cam, R. J. Maraia, and S. I. Grewal. 2006. A role for TFIIIC transcription factor complex in genome organization. Cell 125859-872. [PubMed]
43. Oki, M., and R. T. Kamakaka. 2005. Barrier function at HMR. Mol. Cell 19707-716. [PubMed]
44. Parnell, T. J., J. T. Huff, and B. R. Cairns. 2008. RSC regulates nucleosome positioning at Pol II genes and density at Pol III genes. EMBO J. 27100-110. [PMC free article] [PubMed]
45. Partridge, J. F., B. Borgstrom, and R. C. Allshire. 2000. Distinct protein interaction domains and protein spreading in a complex centromere. Genes Dev. 14783-791. [PMC free article] [PubMed]
46. Paule, M. R., and R. J. White. 2000. Survey and summary: transcription by RNA polymerases I and III. Nucleic Acids Res. 281283-1298. [PMC free article] [PubMed]
47. Roberts, D. N., A. J. Stewart, J. T. Huff, and B. R. Cairns. 2003. The RNA polymerase III transcriptome revealed by genome-wide localization and activity-occupancy relationships. Proc. Natl. Acad. Sci. USA 10014695-14700. [PMC free article] [PubMed]
48. Rusche, L. N., A. L. Kirchmaier, and J. Rine. 2002. Ordered nucleation and spreading of silenced chromatin in Saccharomyces cerevisiae. Mol. Biol. Cell 132207-2222. [PMC free article] [PubMed]
49. Scott, K. C., S. L. Merrett, and H. F. Willard. 2006. A heterochromatin barrier partitions the fission yeast centromere into discrete chromatin domains. Curr. Biol. 16119-129. [PubMed]
50. Scott, K. C., C. V. White, and H. F. Willard. 2007. An RNA polymerase III-dependent heterochromatin barrier at fission yeast centromere 1. PLoS ONE 2e1099. [PMC free article] [PubMed]
51. Simms, T. A., E. C. Miller, N. P. Buisson, N. Jambunathan, and D. Donze. 2004. The Saccharomyces cerevisiae TRT2 tRNAThr gene upstream of STE6 is a barrier to repression in MATalpha cells and exerts a potential tRNA position effect in MATa cells. Nucleic Acids Res. 325206-5213. [PMC free article] [PubMed]
52. Siol, O., M. Boutliliss, T. Chung, G. Glockner, T. Dingermann, and T. Winckler. 2006. Role of RNA polymerase III transcription factors in the selection of integration sites by the Dictyostelium non-long terminal repeat retrotransposon TRE5-A. Mol. Cell. Biol. 268242-8251. [PMC free article] [PubMed]
53. Soutourina, J., V. Bordas-Le Floch, G. Gendrel, A. Flores, C. Ducrot, H. Dumay-Odelot, P. Soularue, F. Navarro, B. R. Cairns, O. Lefebvre, and M. Werner. 2006. Rsc4 connects the chromatin remodeler RSC to RNA polymerases. Mol. Cell. Biol. 264920-4933. [PMC free article] [PubMed]
54. Umylny, B., G. Presting, and W. S. Ward. 2007. Evidence of Alu and B1 expression in dbEST. Arch. Androl. 53207-218. [PubMed]
55. Wallace, J. A., and G. Felsenfeld. 2007. We gather together: insulators and genome organization. Curr. Opin. Genet. Dev. 17400-407. [PMC free article] [PubMed]
56. Wang, L., R. A. Haeusler, P. D. Good, M. Thompson, S. Nagar, and D. R. Engelke. 2005. Silencing near tRNA genes requires nucleolar localization. J. Biol. Chem. 2808637-8639. [PMC free article] [PubMed]
57. Warner, J. R. 1999. The economics of ribosome biosynthesis in yeast. Trends Biochem. Sci. 24437-440. [PubMed]
58. West, A. G., M. Gaszner, and G. Felsenfeld. 2002. Insulators: many functions, many mechanisms. Genes Dev. 16271-288. [PubMed]
59. Yu, Q., J. Sandmeier, H. Xu, Y. Zou, and X. Bi. 2006. Mechanism of the long range anti-silencing function of targeted histone acetyltransferases in yeast. J. Biol. Chem. 2813980-3988. [PubMed]

Articles from Eukaryotic Cell are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...